High-Resolution Spectroscopic Search for the Thermal Emission of the Extrasolar Planet HD 217107 B
Total Page:16
File Type:pdf, Size:1020Kb
A&A 529, A88 (2011) Astronomy DOI: 10.1051/0004-6361/201015802 & c ESO 2011 Astrophysics High-resolution spectroscopic search for the thermal emission of the extrasolar planet HD 217107 b P. E. Cubillos1,2,P.Rojo1, and J. J. Fortney3 1 Department of Astronomy, Universidad de Chile, Santiago, Chile e-mail: [email protected] 2 Department of Physics, Planetary Sciences Group, University of Central Florida, FL 32817-2385, USA 3 Department of Astronomy and Astrophysics, University of California, Santa Cruz,CA 95064, USA Received 22 September 2010 / Accepted 28 February 2011 ABSTRACT We analyzed the combined near-infrared spectrum of a star-planet system with thermal emission atmospheric models, based on the composition and physical parameters of the system. The main objective of this work is to obtain the inclination of the orbit, the mass of the exoplanet, and the planet-to-star flux ratio. We present the results of our routines on the planetary system HD 217107, which was observed with the high-resolution spectrograph Phoenix at 2.14 μm. We revisited and tuned a correlation method to directly search for the high-resolution signature of a known non-transiting extrasolar planet. We could not detect the planet with our current data, but we present sensitivity estimates of our method and the respective constraints on the planetary parameters. With a confidence level of 3-σ we constrain the HD 217107 b planet-to-star flux ratio to be less than 5 × 10−3. We also carried out simulations on other planet candidates to assess the detectability limit of atmospheric water on realistically simulated data sets for this instrument, and we outline an optimized observational and selection strategy to increase future probabilities of success by considering the optimal observing conditions and the most suitable candidates. Key words. planets and satellites: detection – techniques: spectroscopic – stars: individual: HD 217107 1. Introduction starlight reflected from the giant exoplanet Tau-Boötis b, they found an upper limit to the albedo and radius using a least- The characterization of the over 500 detected exoplanets has squares deconvolution method that is well described in the ap- now begun to take place. Most of the studies are carried out at pendices of Collier Cameron et al. (2002), later the author re- optical and infrared wavelengths, because this is where the plan- peated the analysis on υ Andromedab (Collier Cameron et al. etary reflected light and thermal emission peak, respectively. The 2002). Recently Rodler et al. (2008, 2010) searched in the vis- discovery of transiting planets (Charbonneau et al. 2000; Henry ible spectra of HD 75289Ab and Tau-Boötis b and found upper et al. 2000) allowed astronomers to constrain new physical pa- limits for their albedos using a model synthesis method. They rameters such as the radii and masses of the planets, which are constructed a model of the observation composed by a stellar not measurable by the radial velocity method alone. It is on these template plus a shifted and scaled-down version of the stel- systems that in the last years the planetary atmosphere charac- lar template to simulate the starlight reflected from the planet, terization has achieved the most exciting progress through the these models were compared to the data by means of χ2.Inthe use of spectroscopy and broadband photometry with space tele- near-infrared, several attempts have been made to detect Hot- scopes. Examples are the identification of molecules such as wa- Jupiters by trying to distinguish the planetary thermal emission ter absorption (e.g. Tinetti et al. 2007) or methane (Swain et al. from the starlight (Wiedemann et al. 2001; Lucas & Roche 2002; 2008), or the observation of the thermal emission variation with Barnes et al. 2007, 2008, 2010), they also found upper limits orbital phase (Knutson et al. 2007). for the emitted flux of the planets. All these authors have used Although great improvements in characterizing the compo- their own variation of a method based on the same principle sition of transiting Hot-Jupiters have been achieved, they only 1 of separating the planetary and stellar spectra given their rela- represent about 20% of the known extrasolar planets . The char- tive Doppler shifts. Only recently, Snellen et al. (2010) claimed acterization of non transiting planets would require the direct the detection of carbon monoxide from the transmission spec- detection of their light, but the very low flux ratios between the trum of HD 209458 b during a transit observation by using high- planets and their host stars makes a direct detection a very chal- resolution spectra; nonetheless, his technique required a transit- lenging goal. Secondary eclipse observations from Spitzer show ing system. that planet-to-star flux ratios can be as high as 2.5 × 10−3 be- tween 3.6 and 24 μm(e.g.Knutson et al. 2008). At 2.14 μmthe In this work we present an effort to constrain new physical expected flux should be less than these values. Many authors parameters of the non-transiting Hot-Jupiter HD 217107 b. We have attempted a direct detection of the Doppler-shifted sig- attempt to trace its Doppler-shifted signature (estimated to be −4 nature in high-resolution spectroscopy from ground-based tele- ∼10 times dimmer than the star flux) with a correlation func- scopes. In the optical Cameron et al. (1999) tried to observe the tion between high-resolution data and models of its atmospheric spectrum. With positive detections this method would provide 1 www.exoplanet.eu new information on its characteristics, such as its temperature, Article published by EDP Sciences A88, page 1 of 7 A&A 529, A88 (2011) Table 1. Orbital parameters of HD 217107. Parameter Value References Star: Spectral type G8 IV W07 s Teff(K) 5 646 ± 26 W07 K (mag) 4.536 ± 0.021 C03 d (pc) 19.72 ± 0.30 P97 Ms (M)1.02± 0.05 S04 −1 Ks (m s ) 140.6 ± 0.7 W07 −1 vg (km s ) −14.0 ± 0.6 N04 Planet: P (days) 7.12689 ± 0.00005 W07 Tp (JD) 2 449 998.50 ± 0.04 W07 e 0.132 ± 0.005 W07 Orbital Phase m sin i (M )1.33± 0.05 W07 p Jup Fig. 1. Radial velocity curve of HD 217107 vs. orbital phase. The a (AU) 0.074 ± 0.001 W07 crosses mark the observations of Wittenmyer et al. (2007), which we ω (deg) 22.7 ± 2.0 W07 used to compute this orbital solution. The boxes over the curve indicate References. W07: Wittenmyer et al. (2007); C03: Cutri et al. (2003); the coverage of our observations, the filled boxes represent the runs uti- P97: Perryman & ESA (1997); S04: Santos et al. (2004). lized in the analysis, while the open boxes represent the discarded runs (details in Sect. 4.2). chemical composition, and the presence of chemical tracers as- sociated with life. At the same time, the method enables the cali- Tp, P, e,ω,and Ks); and the inclination of the orbit, i,andalso bration of high-resolution spectroscopic models for a larger sam- on the velocity of the center of mass of the system, vg, when mea- ple of planets that do not necessarily transit their parent star. sured from Earth. Thus, the radial velocity curve of the planet In Sect. 2 we review the planetary system HD 217107; in is a distinctive curve in time, parameterized by the values sum- Sect. 3 we describe the observations, data reduction, and calibra- marized in Table 1, where the only unknown parameter is the tion procedures; in Sect. 4 we detail the theoretical atmospheric inclination of the orbit. Figure 1 shows the radial velocity curve spectrum of the planet and the method used to extract and ana- of the star owing to the interaction with HD 217107 b, phased lyze the planetary signal and present the results of our data; in over one orbit, with the origin in phase (φ = 0) at the time of pe- Sect. 5 we develop a strategy for the ideal data acquisition situ- riastron. The radial velocity of the planet is proportional to this ation and simulate observations of other planetary systems; and radial velocity curve (Eq. (1)). in Sect. 6 we give the conclusions of our work. 2.3. Flux estimate 2. The planetary system HD 217107 By simulating the spectra of the planet and its host star as black 2.1. HD 217107 b discovery bodies, we can estimate the order of magnitude of the planet- to-star flux ratio as a function of wavelength. The black body HD 217107 is a main-sequence star that is similar to the Sun emission, F (T), is determined by the surface temperature of the ff λ in mass, radius, and e ective temperature; its spectral type, G8 object. While for the star the temperature is well known from IV, indicates that it is starting to evolve into the red-giant phase models (see Table 1), for the planet our best approximation is (Wittenmyer et al. 2007). The presence of HD 217107 b was first the equilibrium temperature reported by Fischer et al. (1999) through radial velocity mea- surements of the star, the detection was then confirmed by Naef − 1/4 1/2 et al. (2001). Later, Fischer et al. (2001) identified a trend in 1 A Rs Teq = Teff. (2) the residuals of the fit, and Vogt et al. (2005) postulated the ex- 4 a istence of a third companion in an external orbit with a period of 8.6 ± 2.7 yr. The presence of this third object promoted the For a reference value of the bond albedo of A = 0, we found an study of this system in subsequent surveys (Butler et al.